Prominent Chinese scientist Zhu Songchun has called for moving beyond superficial achievements in artificial intelligence in China and focusing on its theoretical and philosophical principles, stressing the importance of forging an independent Chinese approach to ensure genuine leadership in this field.
Article contents:
- Introduction: The necessity of deep understanding
- The five layers of innovation
- Artificial General Intelligence (AGI)
- Challenges in the Chinese market
- Future applications
- Conclusion
- Frequently Asked Questions
Introduction: The necessity of deep understanding
The call by Chinese scientist Zhu Songchun, director of the Beijing Institute for General Artificial Intelligence, has emerged regarding the need to move past the current state of intensive investment in superficial AI applications and instead focus on its theoretical and philosophical foundations. These remarks were made at the Zhongguancun Forum in Beijing, where Zhu described current so-called “achievements” as resembling the tip of an iceberg, concealing deep levels beneath that require serious study.
The five layers of innovation
Zhu compared the stages of AI innovation in China to an iceberg consisting of five layers: implementation, algorithms, models, theory, and philosophy. He pointed out that most companies focus on the top three layers above the water’s surface while neglecting deeper theories and philosophy. Zhu believes that neglecting these layers could limit China’s ability to make breakthrough innovations and render it dependent on a Western approach ill-suited to the local context.
Artificial General Intelligence (AGI)
Zhu touched upon the issue of Artificial General Intelligence, which aims to emulate human cognitive and creative capabilities. He warned against relying solely on massive amounts of data and computing power, emphasizing that “giving machines a heart” means augmenting them with values and ethics that support balanced decision-making. Zhu maintains that China’s role in this framework should focus on building causal and value-based rules to ensure the prevention of dangerous deviations.
Challenges in the Chinese market
The local market is teeming with startups trying to impress investors with quick results or eye-catching applications, but Zhu believes in the necessity of developing a research environment free from time and financial pressures. Achieving quick technical feats is insufficient without establishing theoretical foundations that ensure sustainable progress, especially in the face of competition from global companies such as OpenAI and DeepMind.
Future applications
1. Economic policies: AI built on theoretical and philosophical foundations can contribute to more accurate economic simulations, helping decision-makers formulate long-term plans.
2. Social fields: It could be used in social interaction models or to develop educational curricula compatible with value-driven AI.
3. Scientific research: It could enhance analytical capabilities and the extraction of findings in fields such as medicine, climate science, and others.
Conclusion
Zhu Songchun’s vision highlights the urgent need to rethink prevailing strategies for AI development. One should not settle merely for software achievements and the ability to train models; rather, it is necessary to go further to forge robust foundations that ensure the sustainable and ethical use of these technologies. This approach would position China for genuine leadership rather than temporary superiority.
Frequently Asked Questions
1. Who is Zhu Songchun?
A Chinese scientist and director of the Beijing Institute for General Artificial Intelligence who advocates focusing on the theoretical and philosophical aspects of AI.
2. What is meant by the AI iceberg?
It refers to the five layers of innovation: implementation, algorithms, models, theory, and philosophy, where the latter remain unseen or neglected.
3. What is the significance of “giving machines a heart”?
It means introducing values and ethics into the architecture of artificial intelligence to maintain decision balance and avoid deviations.
4. How does the Chinese approach differ from the Western one?
Zhu believes that the Western approach focuses on computational scale and data, whereas China needs an approach grounded in a deeper understanding of causal and value factors.
5. Is the Chinese market ready for this vision?
It faces challenges due to the race for quick results, but there is growing awareness of the importance of long-term research.
